American Journal of Epidemiology
Copyright © 2000 by The Johns Hopkins University School of Hygiene and PubOc Health
All rights reserved
Vol. 151, No. 11
Printed in USA.
Exposure to Electromagnetic Fields from Use of Electric Blankets and Other
In-Home Electrical Appliances and Breast Cancer Risk
Tongzhang Zheng,1 Theodore R. Holford,1 Susan Taylor Mayne,1 Patricia Hansen Owens,1 Bing Zhang,2 Peter
Boyle,3 Darryl Carter,1 Barbara Ward,1 Yawei Zhang,1 and Shelia Hoar Zahm4
Exposure to electromagnetic fields (EMFs) from use of electric blankets and other in-home electrical
appliances has been hypothesized to increase breast cancer risk. To test the hypothesis, the authors analyzed
data from a case-control study of female breast cancer conducted in Connecticut in 1994-1997. A total of 608
incident breast cancer patients and 609 age frequency-matched controls, 31-85 years old, were interviewed by
trained study interviewers using a standardized, structured questionnaire to obtain information on lifetime use of
various in-home electrical appliances. A total of 40% of the cases and 43% of the controls reported regular use
of electric blankets in their lifetime, which gave an adjusted odds ratio of 0.9 (95% confidence interval (Cl): 0.7,
1.1). For those who reported using electric blankets continuously throughout the night, the adjusted odds ratio
was 0.9 (95% Cl: 0.7, 1.2) when compared with never users. The risk did not vary according to age at first use,
duration of use, or menopausal and estrogen receptor status. The authors also did not find an association
between use of other major in-home electrical appliances and breast cancer risk. In conclusion, exposure to
EMFs from in-home electrical appliance use was not found to increase breast cancer risk in this study. Am J
Epidemiol 2000;151:1103-11.
breast neoplasms; case-control studies; electromagnetic fields
Chronic exposure to electromagnetic fields (EMFs)
in the range of 50-60 Hz has been hypothesized to
increase female breast cancer risk (1-3). In-home electrical appliances often provide the highest magnetic
field levels experienced in the residence, with electric
blankets being among the largest sources. Some studies have reported that pineal gland exposure to EMFs
from electric blanket use could be 10-40 times greater
than exposure to EMFs associated with electrical
wiring in or around the home (4-6). Several mechanisms have been suggested to explain the potential
relation between EMF exposure and breast cancer risk,
including the following: 1) EMFs reduce the pineal
gland's production of melatonin, a hormone that
inhibits mammary carcinogenesis (1,2, 7-9); 2) EMFs
have a direct effect on circulating prolactin and estrogen (1, 10, 11); or 3) EMFs cause alteration of mitotic
processes (12) and dysfunction of cell membranes
(13). It has also been suggested that EMFs might
reverse the oncostatic effect of melatonin at the site of
action in breast tissue (14).
Limited epidemiologic studies investigating electric
blanket use and breast cancer risk have produced
inconclusive results (15-19). A case-control study
from upstate New York found a borderline significantly increased risk of breast cancer for postmenopausal women (odds ratio (OR) = 1.5,95 percent
confidence interval (Cl): 1.0, 2.2) and for premenopausal women (OR = 1.4, 95 percent Cl: 0.9,
2.2) among those who reported using electric blankets
continuously throughout the night compared with
never users (15, 16). When pre- and postmenopausal
women were combined (17), the odds ratio was 1.5 (95
percent Cl: 1.1, 1.9). Two more recent studies, however, found no increased risk associated with electric
blanket use (18, 19). The study by Gammon et al. (18)
found an odds ratio of 1.0 (95 percent Cl: 0.9, 1.2) for
women under age 45 years and an odds ratio of 1.1 (95
percent Cl: 0.9, 1.4) for those 45 years and over among
women who reported using electric blankets, mattress
pads, or heated water beds when compared with never
users. Another study by Coogan and Aschengrau (19)
reported an odds ratio of 1.0 (95 percent Cl: 0.7, 1.4)
among women who reported using an electric blanket
compared with women who did not.
Received for publication March 18, 1999, and accepted for publication August 5, 1999.
Abbreviations: Cl, confidence interval; EMF, electromagnetic field;
OR, odds ratio.
1
Yale University School of Medicine and Yale Cancer Center, New
Haven, CT.
2
McGill University, Montreal, Canada
3
European Institute of Oncology, Milan, Italy.
4
National Cancer Institute, Bethesda, MD.
Reprint requests to Dr. Tongzhang Zheng, 129 Church Street, Suite
700, New Haven, CT 06510 (e-mail: tongzhang.zheng©yale.edu).
1103
1104
Zheng etal.
Besides electric blankets, there are other in-home
electrical appliances that could be significant sources
for domestic exposure to EMFs (20). These include
electric hair dryers, curling irons, vacuum cleaners,
and electric irons. A potential relation between exposure to these in-home electrical appliances and breast
cancer risk has been suggested, but no study has been
conducted to actually assess the relation. Because of
the inconclusive nature of the association and the
widespread use of the aforementioned appliances, we
examined the relation between in-home use of electrical appliances and the risk of breast cancer in a casecontrol study conducted in Connecticut between 1994
and 1997.
MATERIALS AND METHODS
Study population
Cases were histologically confirmed, incident breast
cancer patients (International Classification of
Diseases for Oncology codes 174.0-174.9) who either
had breast-related surgery at the Yale-New Haven
Hospital, New Haven County, or who were residents
of Tolland County, Connecticut, between January 1,
1994, and December 31, 1997. Subjects were
restricted to 31- to 85-year-old women who had no
previous diagnosis of cancer, with the exception of
nonmelanoma skin cancer, and who were alive at the
time of interview.
Potentially eligible cases from the Yale-New Haven
Hospital were identified using computerized patient
information from the Yale-New Haven Hospital
Surgical Pathology Department, where records of all
newly performed breast-related surgeries are kept. We
consecutively selected all breast cancer patients who
met the study eligibility requirements as described
above. A total of 561 incident breast cancer cases were
identified from Yale-New Haven Hospital, with 432 of
them (77 percent) completing in-person interviews.
From the computerized files, we also randomly
selected 569 potential control patients who had had
breast-related surgery and who were histologically
diagnosed with normal tissue or benign breast diseases
(excluding atypical hyperplasia). Of these, 404 (71
percent) participated in the study. Efforts were made to
frequency match the cases and controls by age (within
5-year intervals) using a 1:1 ratio by adjusting the
number of controls randomly selected in each age stratum every few months.
In addition to the cases and controls recruited from
the Yale-New Haven Hospital, we also recruited cases
and controls from Tolland County, Connecticut.
Tolland and New Haven counties have similar breast
cancer incidence rates, and in recent years they have
had similar breast cancer mortality rates as well. Newly
diagnosed cases with Tolland County addresses were
identified from area hospital records by the Rapid Case
Ascertainment Shared Resource of the Yale
Comprehensive Cancer Center, where the staff are
assigned geographically to survey all of the state's nonpediatric hospitals in order to identify newly diagnosed
cases. Cases identified in the field are sent regularly to
the Rapid Case Ascertainment Shared Resource data
entry staff where the case's demographic information is
entered, verified, and screened against the Connecticut
Tumor Registry database. Connecticut also has reciprocal reporting of cancer cases with adjacent states, facilitating complete ascertainment.
A total of 238 such cases were identified for this
study with 176 of them (74 percent) completing inperson interviews. Population-based controls with
Tolland County addresses were recruited using either
random digit dialing methods for those below age 65 as
described by Hartge et al. (21) or Health Care Finance
Administration files for those aged 65 years and over.
The participation rate from random digit dialingselected controls was 64 percent including the initial
telephone screening and from Health Care Finance
Administration controls, 54 percent. A total of 152 random digit dialing-selected controls and 53 Health Care
Finance Administration controls participated in the
study. Efforts were made to frequency match the cases
and controls by age (within 5-year intervals) using a 1:1
ratio by adjusting the number of controls randomly
selected in each age stratum every few months.
The study pathologist (D. C.) reviewed all the pathologic information for the breast cancer cases and for
the benign breast disease controls recruited from YaleNew Haven Hospital. He also reviewed the pathologic
information for the cases recruited from Tolland
County. Carcinomas were classified as in situ, invasive
ductal, or invasive lobular and were staged according
to the tumor-nodes-metastasis system (22).
Interviews
All procedures were performed in accordance with a
protocol approved by the Yale Human Investigations
Committee. After approval by the hospitals and by
each subject's physician, or following selection
through random sampling, potential participants were
approached by letter and then by phone. Those who
agreed were interviewed by trained study interviewers,
either at the subject's home or at a convenient location.
A standardized, structured questionnaire was used to
obtain information on electric blanket usage. The
respondents were asked whether they had used an electric blanket regularly at any time in their life. If their
response was affirmative, the respondents were then
Am J Epidemiol Vol. 151, No. 11, 2000
Electromagnetic Fields and Breast Cancer Risk
asked to provide information related to each period
when they had used an electric blanket. Specifically,
they were asked to give their age at first/each use, age
when they stopped use, the number of years of use, and
finally, the number of months per year and days per
week during months of reported use. Respondents
were also questioned about their most typical pattern
(or mode) of use, that is, whether their electric blanket
was left on for most of/the entire night, if it was only
used to warm the bed, or if it was left on for up to 1
hour or 1-3 hours after getting into bed. Most of the
participants in this study reported the same mode of
use over the different periods of use reported. Only
eight controls and one case reported different modes of
use over different periods of use. The results were
essentially the same even when the reported mode of
use spanned a lifetime as opposed to being only the
current mode of use. Therefore, in tables 2 and 3, we
reported only the lifetime mode of use. The nine subjects who reported different modes of use over different periods were grouped into the other use pattern
category in tables 2 and 3.
Information on the use of other in-home electrical
appliances (such as electric hair dryers, electrically
heated waterbeds, and electric heating pads) was also
collected. Briefly, respondents were questioned about
their age at first use, the number of years of use, and
the number of days per month during those months of
reported use.
Information on other potential confounding factors,
including menstrual and reproductive history, family
cancer history, occupation, diet, and demographic factors, was also collected during the interview. Dietary
information was collected using a scannable semiquantitative food frequency questionnaire developed
by the Fred Hutchinson Cancer Research Center and
specifically designed to optimize estimation of fat
intake. Each subject was asked to characterize her
usual diet in the year prior to being interviewed.
Data analysis
The primary analyses involved comparisons of
usage of various in-home electrical appliances
between cases and controls. Since it has been suggested that EMF exposure may differentially affect
premenopausal and postmenopausal breast cancer risk
(16, 23), we stratified data by menopausal status. In
addition, since it has been suggested that breast cancer
risk related to EMF exposure may vary based on hormone receptor status (24), we examined the association by stratifying on estrogen receptor status. About
92 percent (399/432) of the breast cancer patients
recruited from the Yale-New Haven Hospital have
information on estrogen receptor levels. These levels
Am J Epidemiol
Vol. 151, No. 11, 2000
1105
were measured immunohistochemically at the pathology department of Yale-New Haven Hospital and were
considered to be positive with an H-score larger than
75 as described by McCarty et al. (25) (an "H-score"
is calculated as the percentage of stains x the intensity
of the stain). Detailed information also allowed us to
examine breast cancer risk by cumulative exposure
(lifetime duration in months of use), by duration of
exposure (number of years of use), and by intensity of
exposure (such as the average number of months of
use per year and the average number of days of use per
year).
Unconditional logistic regression was used to estimate the association between exposure and disease and
to control for potential confounders. Potential confounding variables included in the final model were
age (<50, >50 years), body mass index (<22, 22-24.9,
25-29.9, >30 kg/m2), age at menarche (<13, 13-15,
>16 years, or unknown), lifetime months of lactation
(0, 1-6, 7-12, £13), age at first full-term pregnancy
(nulliparous, <20, 20-25, £26 years), dietary fat intake
in grams per day (<46, 47-71, >72, or unknown),
income 10 years before disease diagnosis or interview
($<20,000, $20,000-24,999, $£25,000, or unknown),
race (Whites, Blacks, and others), family breast cancer
history in a first degree relative, and study site (YaleNew Haven Hospital or Tolland County, included in all
models excluding those in table 3, which stratified by
study site). Odds ratios and 95 percent confidence
intervals were calculated using SAS statistical software (26). Tests for trend were conducted by using a
likelihood ratio statistic in a logistic regression model.
RESULTS
As shown in table 1, cases were slightly older than
controls despite the attempt at matching. Women with a
later age at first full-term pregnancy showed a significantly increased risk. In comparison with those less
than age 20 at first full-term pregnancy, the odds ratio
was 1.6 (95 percent CI: 1.1, 2.4) for those having first
full-term pregnancy at ages 20-25. The odds ratio was
1.8 (95 percent CI: 1.1,2.7) for those having a first fullterm pregnancy at ages 26 and over. Those having a
lifetime lactation history of more than 12 months experienced a reduced risk (OR = 0.8, 95 percent CI: 0.5,
1.1) compared with those who never lactated. A history
of having a first-degree relative with breast cancer was
associated with a nonsignificantly increased risk of
breast cancer (OR = 1.2, 95 percent CI: 0.9, 1.6).
Dietary fat intake at the second tertile showed a slightly
increased risk. These observations are generally in line
with what is known about breast cancer etiology.
Data relating electric blanket use and breast cancer
risk are presented in table 2. A total of 40 percent of the
1106
Zheng etal.
TABLE 1. Selected characteristics of breast cancer cases and controls, Connecticut, 1994-1997
No. of cases
(n-608)
No. of controls
(n = 609)
OR*,t
95% CI*
186
422
234
375
1.0
1.3
1.0, 1.7
Race
Whites
Blacks
Others
551
38
19
557
35
17
1.0
1.2
1.2
0.7, 2.1
0.6, 2.4
Annual income ($)
<20,000
20,000-24,999
£25,000
Unknown
354
46
102
106
343
60
119
87
1.0
0.7
0.8
1.1
0.4, 1.0
0.6, 1.1
0.8, 1.5
Body mass index (kg/rrr*)
<21.0
21.0-24.9
£25.0
80
238
290
105
234
270
1.0
1.3
1.3
0.9, 1.9
0.9, 1.8
173
230
186
19
207
199
183
20
1.0
1.4
1.2
1.2
1.1, 1.9
0.9, 1.7
0.6, 2.4
67
245
290
6
66
259
283
1
1.0
1.0
1.1
6.1
0.7, 1.2
0.8, 1.6
0.7, 53.6
54
260
209
85
73
240
198
98
1.0
1.6
1.8
1.4
1.1, 2.4
1.1, 2.7
0.8, 2.3
390
96
46
76
366
101
50
92
1.0
0.8
0.9
0.8
0.6, 1.2
0.5, 1.3
0.5, 1.1
463
145
485
124
1.0
1.2
0.9, 1.6
Characteristics
Age (years)
<50
£50
Fat intake (g/day)
<46
46-71
£72
Unknown
Age at menarche (years)
£15
13-14
<13
Unknown
Age at first full pregnancy (years)
<20
20-25
£26
Nulliparous
Lifetime lactation (months)
0
1-6
7-12
£13
Family breast cancer history
No
Yes
* OR, odds ratio; Cl, confidence interval.
t Odds ratios for each selected characteristic were adjusted for all other selected characteristics listed In table 1.
cases and 43 percent of the controls reported regular
use of an electric blanket, yielding an adjusted odds
ratio of 0.9 (95 percent CI: 0.7,1.1). There was no clear
pattern to the association between age of first use and
breast cancer risk. Breast cancer risk did not vary by the
typical pattern of use. The adjusted odds ratio was 0.9
(95 percent CI: 0.7, 1.1) for those who reported using
electric blankets continuously throughout the night
when compared with nonusers. There was also no
excess risk associated with increasing the duration of
electric blanket use. An adjusted odds ratio of 0.8 (95
percent CI: 0.6, 1.1) was observed for those using an
electric blanket for 37 or more months when compared
with nonusers (table 2). The risk of breast cancer also
Am J Epidemiol Vol. 151, No. 11, 2000
Electromagnetic Fields and Breast Cancer Risk
TABLE 2.
1107
Electric blanket use and risk of breast cancer, Connecticut, 1994-1997
Blanket use
Never
Ever
Age (years) at first use
<25
25-34
35-^4
^45
No. of
No. of
controls
367
241
49
55
79
58
OR*,t
95% Cl*
346
263
1.0
0.9
0.7, 1.1
51
81
69
62
0.7
1.1
0.8
237
90
30
25
96
81
42
32
108
p for trend
Mode of use
To warm bed only
Throughout night
Other use patterns
Missing information
0.7,
0.5,
0.8,
0.5,
1.6
1.0
1.6
1.1
0.34
p for trend
Months of lifetime use
1-12
13-24
25-36
1.0
1.1
0.7
0.7
0.8
0.8, 1.6
0.4,1.2
0.4,1.3
0.6,1.1
0.17
56
147
37
1
64
159
39
1
0.8
0.9
1.2
0.6,1.2
0.7, 1.2
0.7, 2.0
* OR, odds ratio; Cl, confidence interval.
t Adjusted for age, body mass index (kg/m2), lifetime months of lactation, age at menarche, age at first fullterm pregnancy, family breast cancer history, fat intake, race, income 10 years before disease diagnosis or interview, and study site. Never use is the referent category for all odds ratios.
was not associated with the number of years of use,
with the average number of months of use per year, or
with the average number of days of use per year (data
not shown). The results stratified by study site did not
show an increased risk of breast cancer associated with
electric blanket use among subjects from either YaleNew Haven Hospital or Tolland County (table 3).
We also explored the possibility that the risk might
be apparent only among current users as has been
shown for other breast cancer risk factors (27). We
compared both the cases and controls among current or
former users with never users, by age at first use, duration of use, and mode of use. We did not find an
increased risk of breast cancer associated with these
variables among either current users or former users.
Among the 115 cases and 147 controls comprising the
current user group, we found an odds ratio of 0.5 (95
percent Cl: 0.2, 1.1) for those whose mode of use was
"throughout night," an odds ratio of 1.2 (95 percent Cl:
0.6, 2.4) for those with mixed use patterns, and an odds
ratio of 0.7 (95 percent Cl: 0.3, 1.5) for those whose
mode of use was to warm the bed only.
Stratification by menopausal status also showed no
increased risk of breast cancer associated with electric
blanket use among either premenopausal {n — 342) or
postmenopausal {n = 875) women. For example, for
those who reported leaving an electric blanket on all
Am J Epidemiol Vol. 151, No. 11, 2000
night, the adjusted odds ratio was 0.8 (95 percent Cl:
0.5, 1.5) for premenopausal women and 0.8 (95 percent Cl: 0.6, 1.2) for postmenopausal women when
compared with never users. Stratification of cases by
estrogen receptor status (n = 220 for estrogen receptor
positive and n = 179 for estrogen receptor negative)
also showed no increased risk associated with age at
first use, duration of use, and typical pattern of electric
blanket use. For example, for those who reported leaving an electric blanket on all night, the adjusted odds
ratio was 1.1 (95 percent Cl: 0.7, 1.7) for estrogen
receptor-positive patients and 0.9 (95 percent Cl: 0.6,
1.5) for estrogen receptor-negative patients.
Table 4 presents the odds ratios for breast cancer risk
and use of other major in-home electrical appliances.
None of the odds ratios was statistically significant,
except for the estimate for using an electric razor for
1-35 months (OR = 1.7, 95 percent Cl: 1.1, 2.8).
Among those who reported using an electric razor for
more than 36 months, however, the risk was borderline
significantly reduced (OR = 0.6, 95 percent Cl: 0.3,
1.0).
DISCUSSION
This study does not support an association between
risk of female breast cancer and exposure to EMFs in
1108
Zheng et aJ.
TABLE 3.
Electric blanket use and risk of breast cancer by study site, Connecticut, 1994-1997
YNHH1•'subjects
Blanket use
Never
Ever
Age (years) at first use
<25
25-34
35-44
£45
Tol land County
No. of
cases
No. of
controls
No. of
cases
No. of
controls
OR*,t
95%CI»
ORt
95% Cl
260
172
237
167
1.0
1.0
0.7,1.3
107
69
109
96
1.0
0.7
0.5,1.1
35
39
60
38
31
57
42
37
1.2
0.7
1.3
0.8
0.7, 2.1
0.5,1.1
0.9, 2.1
0.5,1.3
14
16
19
20
20
24
27
25
0.8
0.5
0.8
0.7
0.4,1.8
0.3,1.2
0.4,1.5
0.4,1.5
64
20
19
69
55
19
20
73
1.2
1.1
0.9
0.8
0.8,1.8
0.6, 2.2
0.4, 1.7
0.6, 1.2
26
10
6
27
26
23
12
35
1.0
0.4
0.5
0.7
0.5,1.9
0.2,1.0
0.2, 1.5
0.4, 1.4
37
110
25
38
98
30
0.9
1.0
1.0
0.6, 1.5
0.7, 1.5
0.6,1.9
19
37
12
1
26
61
9
0
0.7
0.6
1.7
0.3, 1.3
0.4,1.0
0.6, 4.9
0
1
p for trend
Months of lifetime use
1-12
13-24
25-36
237
p for trend
Mode of use
To warm bed only
Throughout night
Other use patterns
Missing information
* YNHH, Yale-New Haven Hospital; OR, odds ratio; Cl, confidence interval.
t Adjusted for age, body mass index (kg/m8), lifetime months of lactation, age at menarche, age at first fullterm pregnancy, family breast cancer history, fat intake, race, and income 10 years before disease diagnosis or
interview. Never use Is the referent category for all odds ratios.
the range of 50-60 Hz from in-home electrical appliance use, including but not limited to electric blankets.
The risk did not vary by menopausal status or by hormone receptor status. These results are in contrast to
the results by Vena et al. (15-17), yet consistent with
two recent case-control studies (18,19), which also did
not find an association between electric blanket use
and breast cancer risk.
Experimental studies have suggested a potential
relation between exposure to EMFs and breast cancer
risk. A study by Liburdy et al. (14) found that a 12-mG,
60-Hz magnetic field can reverse melatonin's oncostatic action on a human MCF7 breast cancer cell line in
culture. Blask (28) suggested that melatonin may be
directly oncostatic, and this has led Stevens (24) to
speculate that lower melatonin production due to EMF
exposure might increase risk by releasing previously
quiescent transformed cells from growth inhibition.
It may be difficult, however, to directly compare the
results from epidemiologic studies with those from
laboratory studies. The exposure characteristics, such
as time of exposure and the levels of exposure, could
be very different between the two types of studies. For
example, the exposure conditions in the experimental
studies are generally well defined, while the exposures
from use of in-home appliances are quite the opposite.
Furthermore, the exposure levels are usually much
lower and vary between in-home appliances. A recent
study by Hatch et al. (29) reported an average level of
magnetic fields of 2.2 \iT from electric blankets that is
lower than what was reported by others (4, 30). Water
bed heaters produce lower magnetic fields than do
electric blankets; however, they are used throughout
the year, with the strength of magnetic fields depending upon the position of the heater (29). Magnetic
fields from hair dryers can range between 0.100 and
70.0 uT within 6 inches (15.24 cm) of the electric
motor (30). Since curling irons contain only a heating
element but no motor, curling irons produce magnetic
fields considerably lower than those of hair dryers
(29). Considering the significant differences in exposure characteristics, caution should be taken when
comparing the results of epidemiologic and experimental studies.
One potential reason that recent epidemiologic studies have not found an association between exposure to
domestic EMFs and breast cancer risk could be due to
a lack of detail in collected information related to
domestic EMF exposure. Regarding the use of electric
blankets, for example, a number of factors may affect
Am J Epidemiol
Vol. 151, No. 11, 2000
Electromagnetic Fields and Breast Cancer Risk
TABLE 4. Lifetime duration (In months) of In-home electric appliance use and risk of female breast
cancer, Connecticut, 1994-1997
No. of cases
No. of controls
OR'.t
95%CI«
Heated waterbed (months)
Never
<72
£73
Missing information
548
26
34
0
524
46
38
1
1.0
0.6
1.0
0.4, 1.0
0.6, 1.7
Hair dryer (months)
Never
1-59
60-236
>236
181
129
167
131
162
146
146
155
1.0
0.9
1.3
1.0
0.6, 1.2
0.9, 1.8
0.7, 1.4
Curling iron (months)
Never
1-58
£59
378
104
126
352
128
129
1.0
0.8
1.0
0.6,1.1
0.8, 1.4
Heating pad (months)
Never
1-13
2:14
550
30
28
556
26
27
1.0
1.2
1.2
0.7, 2.1
0.7, 2.1
Electric razor (months)
Never
1-35
£36
Missing information
532
52
24
0
541
33
34
1
1.0
1.7
0.6
1.1,2.8
0.3, 1.0
Electric toothbrush (months)
Never
1-32
£33
513
38
57
510
48
51
1.0
0.8
1.0
0.5, 1.2
0.7, 1.6
Vacuum cleaner (months)
Never
1-46
47-70
71-109
£110
58
135
127
147
141
52
136
141
136
144
1.0
1.1
0.9
1.0
0.9
0.7, 1.8
0.5, 1.5
0.6,1.6
0.5, 1.4
Popcorn maker (months)
Never
1-9
£10
503
51
54
509
50
50
1.0
1.2
1.2
0.8, 1.8
0.8, 1.9
Computer (months)
Never
1-38
39-90
£91
312
92
115
89
261
115
118
115
1.0
0.7
0.9
0.7
0.5, 1.0
0.6, 1.2
0.5, 0.9
Iron (months)
Never
1-53
54-100
£101
72
170
188
178
63
181
182
183
1.0
0.9
0.9
0.9
0.6, 1.4
0.6,1.4
0.6,1.4
Appliance use
• OR, odds ratio; Cl, confidence interval.
t Adjusted for age, body mass index (kg/m1), lifetime months of lactation, age at menarche, age at first fullterm pregnancy, family breast cancer history, fat intake, race, income 10 years before disease diagnosis or interview, and study site.
Am J Epidemiol
Vol. 151, No. 11, 2000
1109
1110
Zheng et al.
the actual level of exposure to EMFs, including the
following: the type and routing of the heating element;
the controller design; the means of overheat protection; and human use factors such as heat tolerance, frequency, and pattern of use (31). Lack of detailed information about these factors would cause a
misclassification of exposure. While the misclassification is unlikely to be differential, nondifferential misclassification can still cause underestimation of the
association between in-home electrical appliance use
and breast cancer risk. This argument, however, does
not seem to coincide well with the fact that there was
no apparent increased risk for women who started
using electric blankets at a very early age, used them
for a long time, and kept them on continuously
throughout the night.
Another potential explanation for the lack of association observed in our study concerns the use of benign
breast disease patients as part of the control group. If
benign breast diseases were associated with EMF
exposure, a weak association between use of electrical
appliances and breast cancer risk could be obscured.
However, this explanation does not seem to be highly
likely because population-based controls were used for
subjects recruited from Tolland County. As described
earlier, when these two study sites were analyzed separately, we reached essentially the same conclusions as
in the combined analyses.
Another potential limitation of our study is that we
did not collect information on occupational and residential electric wiring exposure to EMFs. However,
the proportion of the female work force employed in
electrical occupations associated with exposure to
EMFs is small, and most of the studies relating such
occupational exposures to breast cancer risk have produced largely negative results (32, 33). Studies have
also suggested that EMF exposure associated with
electrical wiring in or around the home was substantially smaller than that from electric blanket use or
from the use of other in-home electrical appliances (4,
5, 20). Finally, residential exposure to EMFs also has
not been consistently associated with breast cancer risk
(33-36). Therefore, it is unlikely that the observed lack
of association between in-home electrical appliance
use and breast cancer risk can be entirely explained by
confounding from occupational and residential EMF
exposures.
One of the advantages of our study is that we have
collected detailed information on major in-home electrical appliances and not limited our investigation to
electric blanket use. Examination of the relation
between breast cancer risk and use of electrical appliances other than electric blankets is important since
studies have shown that these other in-home electrical
appliances could also significantly contribute to
human EMF exposure (20).
Our study assessed the effect of lifetime use of
major in-home electrical appliances and was not limited to exposures at just one time. It is currently
unknown which exposure (past or current) is more
important if a relation indeed exists between EMFs
and breast cancer risk. It has been suggested (8) that
EMF exposure many years in the past should be
assessed if stem cell turnover has been affected. On the
other hand, if cancer cell growth is the underlying
mechanism for an association between EMFs and
breast cancer risk, then very recent exposures could be
crucial. Savitz et al. (37) pointed out that, if prolonged
exposure is the basis for purported adverse health
effects, the duration of exposure from different sources
would be of paramount importance. These suggestions
emphasize the importance of examining the relation by
age at first regular use and duration of use of various
electric appliances.
In summary, there has been increasing concern that
exposure to EMFs resulting from in-home electrical
appliance use and residential and occupational exposures may increase breast cancer risk, as well as cause
other adverse health effects (1-3). This possibility has
provoked considerable controversy, as summarized by
others (7-9, 33, 38-40), and is far from resolved.
While recent epidemiologic studies, including this
one, do not support an association between exposure to
EMFs from in-home electrical appliances and breast
cancer risk, several possible limitations may have
hampered these findings. One of the potential limitations discussed earlier is a lack of detail in information
collected through in-person interviews related to
domestic EMF exposure. The resultant misclassification of exposure may cause an underestimation of the
association between in-home electrical appliance use
and breast cancer risk. Future epidemiologic studies
would be needed to improve the measurement tools so
that a better estimate of EMF exposure from various
sources over many years could be achieved, but thus
far the evidence for such an association is not strong.
ACKNOWLEDGMENTS
This study is supported by grant CA-62986 from the
National
Cancer
Institute/National
Institute
of
Environmental Health Science.
The authors thank Donna Carrano, Melita Bosnyak,
Heather Hutson, and Sylvia Ullman for their high quality
interviewing. They also are grateful for assistance provided
by the Yale Cancer Center Shared Resources in ascertaining
the study patients from Tolland County. Special thanks are
Am J Epidemiol Vol. 151, No. 11, 2000
Electromagnetic Fields and Breast Cancer Risk
extended to Drs. Kumiko Iwamoto, Gwen Collman, and G.
Iris Obrams at the National Institutes of Health for their support and guidance during the study. The following
Connecticut hospitals have participated in the study: YaleNew Haven Hospital, Hospital of St. Raphael, Hartford
Hospital, St. Francis Hospital and Medical Center, New
Britain General Hospital, Middlesex Hospital, Mt. Sinai
Hospital, Manchester Memorial Hospital, UCONN Health
Center/John Dempsey Hospital, Windham Community
Memorial Hospital, Day Kimball Hospital, Rockville
General Hospital, and Johnson Memorial Hospital.
REFERENCES
1. Stevens RG. Electric power use and breast cancer a hypothesis. Am J Epidemiol 1987;125:556-61.
2. Stevens RG. Electric power, melatonin, and breast cancer. In:
Gupta D, Attanasio A, Reiter RJ, eds. Pineal gland and cancer.
London, UK: Brain Research Promotion, 1988:223-4.
3. Stevens RS, Scott D, Thomas DB, et al. Electric power, pineal
function, and the risk of breast cancer. FASEB J 1992;6:853-60.
4. Florig HK, Hoburg JF. Power-frequency magnetic fields from
electric blankets. Health Phys 1990;58:493-500.
5. Florig HK, Hoburg JF. Magnetic field exposure associated
with electric blankets. In: Proceedings of the 1988 contractors
review. Biological effects from electric and magnetic fields, air
ions, and ion currents associated with high voltage transmission lines, Phoenix, AZ, October 30-November 3, 1988.
6. Preston-Martin S, Peters JM, Yu MC, et al. Myelogenous
leukemia and electric blanket use. Bioelectromagnetics 1988;
9:207-13.
7. Preston-Martin S. Breast cancer and magnetic fields.
(Editorial). Epidemiology 1996;7:457-8.
8. Stevens RG. Biologically based epidemiological studies of
electric power and cancer. Environ Health Perspect 1993; 101
(suppl 4):93-100.
9. Stevens RG, Davis S. The melatonin hypothesis: electric
power and breast cancer. Environ Health Perspect 1996; 104
(suppl 1): 135-40.
10. Cohen M, Lipman M, Chabner B. Role of pineal gland in aetiology and treatment of breast cancer. Lancet 1978;2:814-16.
11. Stevens RG, Wilson BW, Anderson LE. The question of cancer. In: Wilson BW, Stevens RG, Anderson LE, eds. Extremely
low frequency electromagnetic fields: the question of cancer.
Columbus, OH: Battelle Press, 1990:361-70.
12. Adey WR. Electromagnetic fields, cell membrane amplification, and cancer promotion. In: Wilson BW, Stevens RG,
Anderson LE, eds. Extremely low frequency electromagnetic
fields: the question of cancer. Columbus, OH: Battelle Press,
1990:211-49.
13. Easterly CE. Cancer link to magnetic field exposure: a hypothesis. Am J Epidemiol 1981;114:169-74.
14. Liburdy RP, Sloma TR, Sokolic R. ELF magnetic fields,
breast cancer, and melatonin: 60 Hz fields block melatonin's
oncostatic action on ER+ breast cancer cell proliferation. J
Pineal Res 1993; 14:89-97.
15. Vena JE, Graham S, Hellmann R, et al. Use of electric blankets
and risk of postmenopausal breast cancer. Am J Epidemiol
1991; 134:180-5.
16. Vena JE, Freudenheim JL, Marshall JR, et al. Risk of premenopausal breast cancer and use of electric blankets. Am J
Epidemiol 1994; 140:974-9.
17. Vena JE, Marshall JR, Freudenheim JL, et al. The authors
reply. (Letter). Am J Epidemiol 1995; 142:446-7.
18. Gammon MD, Schoenberg JB, Britton JA, et al. Electric blanket use and breast cancer risk among younger women. Am J
Am J Epidemiol
Vol. 151, No. 11, 2000
1111
Epidemiol 1998;148:556-63.
19. Coogan PF, Aschengrau A. Exposure to power frequency magnetic fields and risk of breast cancer in the Upper Cape Cod
Cancer Incidence Study. Arch Environ Health 1998^3:359-67.
20. Delpizzo V. A model to assess personal exposure to ELF magnetic fields from common household sources. Bioelectromagnetics 1990; 11:139-47.
21. Hartge P, Brinton LA, Rosenthal JF. Random digit dialing
selecting a population-based control group. Am J Epidemiol
1984; 120:825-33.
22. Beahrs OH, Myers MH, eds. Manual for staging cancer. 2nd
ed. Philadelphia, PA: JB Lippincott, 1983:127-34.
23. Wertheimer N, Leeper E. Magnetic field exposure related to
cancer subtypes. Ann N Y Acad Sci 1987;502:43-53.
24. Stevens RG. Re: "Risk of premenopausal breast cancer and use
of electric blankets." (Letter). Am J Epidemiol 1995; 142:446.
25. McCarty KS, Miller S, Cox EB, et al. Estrogen receptor analyses. Arch Pathol Lab Med 1985;109:716-21.
26. SAS Institute, Inc. SAS/STAT user's guide. Version 6. Cary,
NC: SAS Institute, Inc, 1990.
27. Collaborative Group on Hormonal Factors in Breast Cancer.
Breast cancer and hormonal contraceptives: collaborative
reanalysis of individual data on 53,279 women with breast
cancer and 100,239 women without breast cancer from 54 epidemiological studies. Lancet 1996;347:1713-27.
28. Blask DE. The emerging role of the pineal gland and melatonin
in oncogenesis. In: Wilson BW, Stevens RG, Anderson LE,
eds. Extremely low frequency electromagnetic fields: the
question of cancer. Columbus, OH: Battelle Press, 1990:
319-35.
29. Hatch EE, Linet MS, Kleinerman RA, et al. Association
between childhood acute lymphoblastic leukemia and use of
electrical appliances during pregnancy and childhood.
Epidemiology 1998;9:234-45.
30. National Institute of Environmental Health Sciences/
Department of Energy. Questions and answers about EMF;
electric and magnetic fields associated with the use of electric
power. Washington, DC: US Government Printing Office,
1995. (US Government document DOE/EE-0040).
31. Florig HK. Magnetic fields from electric appliances. Presented
at the EPRI Utility Seminar power-frequency electric and
magnetic field exposure assessment, Colorado Springs,
Colorado, October 12-14,1988. Palo Alto, CA: Electric Power
Research Institute, 1988.
32. Goldberg MS, Labreche F. Occupational risk factors for
female breast cancer: a review. Occup Environ Med 1996;53:
145-56.
33. Trichopoulos D. Are electric or magnetic fields affecting mortality from breast cancer in women? (Editorial). J Natl Cancer
Inst 1994;86:885-6.
34. Kheifets LI, Kelsey JL. Epidemiologic studies of electric and
magnetic fields and cancer. In: Lin JC, ed. Advances in electromagnetic fields in living systems. Vol 2. New York, NY:
Plenum Press, 1977.
35. Li CY, Theriault G, Lin RS. Epidemiological appraisal of studies of residential exposure to power frequency magnetic fields
and adult cancers. Occup Environ Med 1996;53:505-10.
36. Polk C, Postow E, eds. Handbook of biological effects of electromagnetic fields. Boca Raton, FL: CRC Press, Inc, 1996.
37. Savitz DA, Pearce NE, Poole C. Methodological issues in the
epidemiology of electromagnetic fields and cancer. Epidemiol
Rev 1989; 11:59-78.
38. Siemiatycki J. Problems and priorities in epidemiologic
research on human health effects related to wiring code and
electric and magnetic fields. Environ Health Perspect 1993;
l01(suppl4):135-41.
39. Stevens RS, Scott D. The melatonin hypothesis: electric power
and breast cancer. Environ Health Perspect 1996;104(suppl 1):
135-40.
40. Shaw GM, Croen LA. Human adverse reproductive outcomes
and electromagnetic field exposures: review of epidemiologic
studies. Environ Health Perspect 1993; 101 (suppl 4):107-19.